aiqbench analysis
TabFM introduces a 400M-parameter foundation model specifically designed for tabular data, leveraging synthetic tables generated from structural causal models to achieve zero-shot performance across diverse benchmarks. By formulating supervised prediction as an in-context learning task, the approach eliminates the need for per-dataset tuning and outperforms traditional AutoML pipelines without requiring specific training on target datasets. This advancement addresses a significant gap where tabular machine learning has historically relied on fragmented workflows rather than generalized foundation models.
TabFM: A Zero-Shot Foundation Model for Tabular Data
arXiv:2609.37959v1 Announce Type: new Abstract: Tabular machine learning typically relies on per-dataset workflows, fitting tree ensembles or running AutoML searches from scratch for every task. We present TabFM, a 400M-parameter tabular foundation model that formulates supervised tabular prediction as in-context learning. TabFM produces calibrated zero-shot predictions in a single forward pass without task-specific tuning. Trained entirely on synthetic tables generated from structural causal models, TabFM learns general tabular representations that transfer zero-shot to real-world tasks. Across all 51 benchmark datasets in TabArena (38 classification and 13 regression), zero-shot TabFM ranks first among default tabular foundation models and outperforms tuned AutoML pipelines. Two extensions over the same frozen weights improve performance further on both tracks: multi-view feature expansion with ensembling and post-hoc calibration (TabFM+), and LLM-guided, dataset-specific data processing and feature engineering (TabFM-Auto).
Abstract:Tabular machine learning typically relies on per-dataset workflows, fitting tree ensembles or running AutoML searches from scratch for every task. We present TabFM, a 400M-parameter tabular foundation model that formulates supervised tabular prediction as in-context learning. TabFM produces calibrated zero-shot predictions in a single forward pass without task-specific tuning. Trained entirely on synthetic tables generated from structural causal models, TabFM learns general tabular representations that transfer zero-shot to real-world tasks. Across all 51 benchmark datasets in TabArena (38 classification and 13 regression), zero-shot TabFM ranks first among default tabular foundation models and outperforms tuned AutoML pipelines. Two extensions over the same frozen weights improve performance further on both tracks: multi-view feature expansion with ensembling and post-hoc calibration (TabFM+), and LLM-guided, dataset-specific data processing and feature engineering (TabFM-Auto).
Machine Learning (cs.LG)
arXiv:2609.37959 [cs.LG]
(or arXiv:2609.37959v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.37959 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 221 words
Related coverage
Introducing TabFM: A zero-shot foundation model for tabular data
Google Research · Jun 30, 2026
TabFM-Auto: Self-Evolving Pipelines for Tabular Foundation Models
arXiv · Sep 30, 2026
Adaptation Interfaces for In-Context Tabular Foundation Models in Time-to-Event Prediction
arXiv · Sep 7, 2026
Tabular Foundation Models for Clinical Survival Analysis via Survival-Aware Adaptation
arXiv · Jun 12, 2026